For the English speakers he said he's got a 2050 cluster running 7.3.3 which he uses nagios and cacti to monitor but he's not sure what limits he should be setting as alarms for cifs ops, cpu and cache age.
Bonjour, vous m'excusez mon francias a l'ecrit j'espere!
a mon avis:
CPU = moins de 75%
CIFS LATENCY = moins de 20 milliseconds
CIFS OPS = ca depend le nombre, type et vitesse des disques